gcl-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Gcl-devel] gcl 2.7.0 hopelessly broken in debian


From: Faré
Subject: Re: [Gcl-devel] gcl 2.7.0 hopelessly broken in debian
Date: Tue, 4 May 2010 17:35:43 -0400

Dear Camm,

thanks a lot for your help.

Based on your feedback, I removed clc and tried again. Passed that
point, I found a few more issues.

1- the fasl compiler tries to be clever with packages and apparently
treats defpackage specially. To go around that issue, I specially
defpackage trivial packages on gcl. I tried to write some complex
macrolet that does the right thing on either gcl or other systems, but
somehow it didn't make gcl happy.

2- relative pathnames are parsed by gcl without a :relative in the
directory component. That's a gcl bug. I now work around in
merge-pathnames* and relativize-pathname-directory.

3- GCL borks on this:
(MAKE-PATHNAME :HOST NIL :DEVICE NIL :DIRECTORY
           '(:RELATIVE :WILD-INFERIORS) :DEFAULTS #P"/**/*.*")

4- There seem to be errors in printing conditions that prevent our
print-methods from being called.

5- GCL is still missing support for ~W.

6- something prevents janderson's logical pathname tests from working on GCL.

7- when compiling fare-utils, it looks like ASDF itself is working,
but GCL borks on base/macros.lisp.

Enough for today. I'm releasing ASDF 1.715, which I believe works
better on GCL than any previous version of ASDF. You're welcome to
include it in GCL, submit patches, etc.

[ François-René ÐVB Rideau | Reflection&Cybernethics | http://fare.tunes.org ]
You can tell whether a man is clever by his answers.
You can tell whether a man is wise by his questions. -- Naguib Mahfouz




On 4 May 2010 13:36, Camm Maguire <address@hidden> wrote:
> Greetings, and thank you so much for your reply!
>
> There is also a gcc-4.4 -g -Os bug on gcl_pcl_methods.c, now worked
> around in the newly uploaded -100 package.  The clc stuff now works
> for me.  Please let me know if this is not the case for you too.
>
> If you are interested in gcl development, please let me know, as I
> have a local tree with significant enhancements as yet uncommitted,
> due to lack of time (writing a book).
>
> Take care,
>
> Faré <address@hidden> writes:
>
>> Can you provide an updated gcl installation script?
>>
>> When I simply replace c-l-c: by clc: I get other errors.
>>
>> As for the missing function ASDF:PROCESS-SOURCE-REGISTRY, are you
>> using the latest ASDF? Say 1.704 from debian or 1.714 from upstream?
>>
>> I removed the clc image and tried again. I can compile asdf, but
>> gclcvs won't load it, saying package ASDF does not exist - apparently
>> the eval-when doesn't get a chance to be run before gclcvs tries to
>> link stuff to said package.
>>
>> I also tried loading asdf.lisp as a lisp file then loading a system,
>> but then I get an
>> Error in error:
>> ERROR SIMPLE-TYPE-ERROR (DATUM INTERNAL-SIMPLE-MISSING-COMPONENT
>>                                EXPECTED-TYPE
>>                                (SATISFIES CONDITION-CLASS-P)
>>                                FORMAT-CONTROL Not a condition type: ~S
>>                                FORMAT-ARGUMENTS
>>                                (INTERNAL-SIMPLE-MISSING-COMPONENT)) NIL
>>
>> How do I get a useful backtrace from there?
>>
>> [ François-René ÐVB Rideau | Reflection&Cybernethics | http://fare.tunes.org 
>> ]
>> There are two distinct classes of men in the Nation, those who pay taxes
>> and those who receive and live upon the taxes. -- Thomas Paine
>>
>>
>>
>>
>> On 2 May 2010 13:13, Camm Maguire <address@hidden> wrote:
>>> Greetings!  The pathology was do to the failure in the clc script,
>>> causing the clc image to be saved in an error state.
>>> unregister.... produces a working image, and a replacement clc
>>> installation works fine.  The nickname c-l-c went to clc.
>>>
>>> But now I have the following -- maybe you can suggest a fix:
>>>
>>> GCL_ANSI=t gclcvs -eval '(asdf:load-system :fare-utils)'
>>> GCL (GNU Common Lisp)  2.7.0 ANSI    Apr  9 2010 18:42:13
>>> Source License: LGPL(gcl,gmp,pargcl), GPL(unexec,bfd,xgcl)
>>> Binary License:  GPL due to GPL'ed components: (XGCL READLINE BFD UNEXEC)
>>> Modifications of this banner must retain notice of a compatible license
>>> Dedicated to the memory of W. Schelter
>>>
>>> Use (help) to get some basic information on how to use GCL.
>>>
>>> Temporary directory for compiler files set to /tmp/
>>>
>>> Error:
>>> Fast links are on: do (si::use-fast-links nil) for debugging
>>> Signalled by ASDF:OPERATE.
>>> SIMPLE-ERROR: Can't get template for #<Standard-Method 
>>> ASDF:PROCESS-SOURCE-REGISTRY (SYMBOL) 130356220>
>>>
>>> Broken at ASDF:OPERATE.  Type :H for Help.
>>>  1 (Continue) Return to top level.
>>> COMMON-LISP-USER>>
>>>
>>> Take care,
>>>
>>> Faré <address@hidden> writes:
>>>
>>>> To reproduce the problem, install the latest cl-asdf 1.704 from debian
>>>> and common-lisp-controller 7.2 with gclcvs 2.7.0-64 and try this:
>>>>
>>>> GCL_ANSI=t gclcvs -eval '(asdf:load-system :fare-utils)'
>>>>
>>>> Use any system here - I'm using fare-utils from git because I know it
>>>> and it has no dependencies. Try alexandria or whatever.
>>>>
>>>> Without debian, try downloading the latest asdf from git (currently
>>>> 1.710) and try this:
>>>> gcl
>>>> (require :asdf) ; or (load "/path/to/asdf.lisp")
>>>> (asdf:load-system :fare-utils)
>>>>
>>>> PS: it would be nice if gcl were packaged by clbuild.
>>>>
>>>> The error I get is
>>>> Unrecoverable error: bind stack overflow.
>>>>
>>>> (Note: linux amd64)
>>>>
>>>> Haven't tried 32 bit in a while - would have to resurrect a 32-bit
>>>> machine / installation.
>>>>
>>>> [ François-René ÐVB Rideau | Reflection&Cybernethics | 
>>>> http://fare.tunes.org ]
>>>> Death is only a state of mind.
>>>>
>>>> Only it doesn't leave you much time to think about anything else.
>>>>
>>>>
>>>> On 29 April 2010 10:16, Camm Maguire <address@hidden> wrote:
>>>>> Greetings!  If you could please post some failing command lines I can
>>>>> reproduce, that would be great.
>>>>>
>>>>> Right now, I think the issue is that the c-l-c nickname has been
>>>>> dropped in common-lisp-controller.  I'm assuming this is facility is
>>>>> still the "Debian lisp way":
>>>>>
>>>>> INTERNAL-SIMPLE-ERROR: There is no package with the name C-L-C.
>>>>>
>>>>> This is from a installation script used to setup c-l-c in gclcvs.  If
>>>>> I know what will be permanent, I can fix this and re-release.  But not
>>>>> until the mipsel buildd is finished and migration takes place.  Would
>>>>> like to do the kfreebsd-amd64 by then too.
>>>>>
>>>>> Would appreciate any feedback here as I'm more familiar with the older
>>>>> math lisp code than the ansi cl- stuff.
>>>>>
>>>>> Take care,
>>>>>
>>>>> Faré <address@hidden> writes:
>>>>>
>>>>>> Is the GCL 2.7.0 package for debian maintained?
>>>>>>
>>>>>> I get weird Unrecoverable error: bind stack overflow while trying to
>>>>>> use GCL to compile even the simplest ASDF systems. Where exactly the
>>>>>> overflow happens moves *further* when I try tracing more functions, so
>>>>>> somehow that bind stack grows more slowly with traced functions than
>>>>>> untraced functions. Still, eventually it borks for no obvious good
>>>>>> reason.
>>>>>>
>>>>>> Is that a known bug fixed upstream?
>>>>>>
>>>>>> Can you make GCL work with ASDF? With clbuild? That would be nice.
>>>>>>
>>>>>> Currently, I have to declare GCL support for ASDF and CL-Launch
>>>>>> broken. Haven't been able to test positively for months.
>>>>>>
>>>>>> [ François-René ÐVB Rideau | Reflection&Cybernethics | 
>>>>>> http://fare.tunes.org ]
>>>>>> Corollaries to the Law of Bitur-Camember: The political process destroys 
>>>>>> the
>>>>>> value of all known resources that are up for grabs. The socialist 
>>>>>> process of
>>>>>> systematically denying legitimacy to property rights applies the 
>>>>>> political
>>>>>> process universally and destroys the value of all available resources.
>>>>
>>>>
>>>>
>>>>
>>>
>>> --
>>> Camm Maguire                                       address@hidden
>>> ==========================================================================
>>> "The earth is but one country, and mankind its citizens."  --  Baha'u'llah
>>>
>>
>>
>>
>>
>
> --
> Camm Maguire                                       address@hidden
> ==========================================================================
> "The earth is but one country, and mankind its citizens."  --  Baha'u'llah
>




reply via email to

[Prev in Thread] Current Thread [Next in Thread]